Understanding the EU Deforestation Regulation & the role of geospatial data | Geo for Good 2023

00:59:18
https://www.youtube.com/watch?v=mqiCB9hxeuk

Summary

TLDRThe video covers a panel discussion on the European Union Deforestation Regulation (EUDR), where experts discuss its implications and challenges. The EUDR mandates that commodities sold in the EU must not be linked to deforestation after December 2020. The panelists emphasize the complexities of mapping and verifying supply chains, the need for transparent data and methodologies, and the integration of geospatial data for effective compliance. They highlight the necessity of collaborative efforts from various stakeholders to meet regulatory demands without hampering economic operations.

Takeaways

  • 🌿 The EUDR targets deforestation-free supply chains for goods entering the EU.
  • 🗺️ Mapping and transparency are critical for EUDR compliance.
  • 📉 Deforestation risk assessment is key, focusing on land status post-2020.
  • 🔍 Geospatial data aids in monitoring land use changes and ensuring compliance.
  • 💼 Companies face challenges in integrating and verifying supply chain data.
  • 🌐 The regulation impacts global trade dynamics, especially in commodities.
  • 🛠️ Technology solutions are needed for tracking and transparency.
  • 🤝 Collaboration among stakeholders is crucial for effective implementation.
  • 📊 Transparent methodologies enhance credibility and compliance.
  • ⚖️ Ethical and economic considerations influence how the regulation is enforced.

Timeline

  • 00:00:00 - 00:05:00

    Alicia Sullivan introduces a panel discussing the EU Deforestation Regulation (EUDR) and its relevance to businesses selling into the EU. Starting in 2024, these companies must demonstrate their products weren't grown on deforested land post-December 2020. The session is divided into expert talks, a panel discussion, and audience questions.

  • 00:05:00 - 00:10:00

    Remi D'Annunzio outlines the compliance process for the EUDR, focusing on the responsibilities of operators to ensure products are deforestation-free and track geolocation data of farms. Discussions also cover regulation texts and FAQs published by the EU Commission to facilitate understanding.

  • 00:10:00 - 00:15:00

    Remi continues by detailing the benchmark system for assessing countries' risk levels: standard, low, or high. This determines the percentage of compliance information to be monitored by EU importing countries—higher percentages for high-risk countries. He also describes ongoing efforts for creating a digital public infrastructure with partners like USAID, WRI, and Google.

  • 00:15:00 - 00:20:00

    Tasso Azevedo emphasizes understanding the scope of application to ensure compliance with EUDR, especially for regions like Brazil's Cerrado where deforestation occurs on non-forest land as defined by the EUDR. He suggests adopting a regional, crop-specific risk classification approach and highlights land use change assessments and traceability challenges.

  • 00:20:00 - 00:25:00

    Tasso continues to explain methods for deforestation assessment and traceability using probabilistic tools. He presents ideas for mass balance control and property-wide assessments to simplify compliance, noting that precise tracking could be impractical.

  • 00:25:00 - 00:30:00

    Pierrick Rambaud of Louis Dreyfus Company describes the immense challenges in mapping supply chains and obtaining precise land geometries needed for EUDR compliance. He underscores the importance of standardization, accurate data gathering, and the high costs associated with using high-resolution imagery for deforestation risk assessment.

  • 00:30:00 - 00:35:00

    Bobby Pinter from NGIS discusses the role of geospatial tools like Tracemark in managing the complexities surrounding EUDR compliance, particularly the aggregation of commodity sources and ensuring deforestation-free certifications amid operational challenges among supply chains such as coffee.

  • 00:35:00 - 00:40:00

    Evan Tachovsky from WRI discusses the importance of digital public infrastructure, tailored applications, and community building to support the EUDR. He stresses the need for open infrastructures for global compliance and preventing unintended negative consequences while maintaining transparency in solutions.

  • 00:40:00 - 00:45:00

    During the panel discussion, experts highlight gaps in compliance, ambiguities in regulations, and the necessity for transparency in methodologies used by commercial providers. They also emphasize the need for property-level assessments and simpler compliance and monitoring strategies to avert deforestation.

  • 00:45:00 - 00:50:00

    Questions from the audience reveal concerns about the feasibility of current compliance strategies and transparency requirements. Panel experts discuss potential simplifications, such as focusing on property rather than plot-based assessments to realistically manage deforestation risks under the EUDR.

  • 00:50:00 - 00:59:18

    Audience concerns also focus on the regulation's approach to transparency and compliance processes, highlighting the need for industry-wide cooperation and clear guidance from regulatory bodies to streamline efforts in recognizing and preventing deforestation effectively.

Show more

Mind Map

Video Q&A

  • What is the EU Deforestation Regulation (EUDR)?

    The EUDR is a regulation set to be implemented starting in 2024, requiring companies that sell goods into the EU to prove that these goods were not grown on land deforested after December 2020.

  • What are the main challenges in complying with the EUDR?

    Main challenges include mapping supply chains accurately, integrating geospatial data for tracking deforestation, and establishing transparent reporting methodologies.

  • What role do commercial data providers play in EUDR compliance?

    Commercial data providers offer tools and data sets for monitoring deforestation and risk assessment, but their methodologies need to be transparent and scientifically robust to be credible.

  • How is deforestation risk determined according to the EUDR?

    Deforestation risk is determined by verifying if the land was forested in 2020 and ensuring that current commodities do not originate from newly deforested land. Precision in geospatial data is crucial.

  • What are "commodities" in the context of EUDR?

    Commodities refer to the main goods like coffee, cocoa, or soybeans that are tracked for potential deforestation impacts according to legislation.

  • What happens if a product is found to be linked to deforestation?

    If a product is linked to deforestation after 2020, it may be prevented from entering the EU market, potentially affecting the entire supply chain from that source.

  • What are the penalties for non-compliance with the EUDR?

    Non-compliance could lead to shipping delays and detainment of goods by EU countries, as they have the right to stop containers for verification.

  • How does the EUDR impact global markets?

    It could significantly alter how supply chains operate, potentially excluding deforestation-linked products from entering European markets and incentivizing sustainable practices.

  • Why is transparency important for EUDR compliance?

    Transparency in methodologies and data used for compliance is crucial to ensure credibility and trust in the reported findings, facilitating smoother regulatory approval processes.

  • What technologies can assist in EUDR compliance?

    Technologies like satellite imagery, blockchain for traceability, and digital platforms for data integration play integral roles in ensuring compliance.

View more video summaries

Get instant access to free YouTube video summaries powered by AI!
Subtitles
en
Auto Scroll:
  • 00:00:00
    [MUSIC PLAYING]
  • 00:00:02
  • 00:00:15
    ALICIA SULLIVAN: All right.
  • 00:00:16
    Hey, everyone, I'm Alicia Sullivan.
  • 00:00:17
    I'm a product manager on Earth Engine.
  • 00:00:19
    And I'm really excited to have a panel, here, today,
  • 00:00:22
    to talk about understanding the EU Deforestation
  • 00:00:25
    Regulation and the Role of Geospatial Data.
  • 00:00:28
    Can you all hear me OK?
  • 00:00:30
    OK, great.
  • 00:00:31
    Awesome.
  • 00:00:33
    So we have an amazing panel, here,
  • 00:00:35
    that are going to talk, here, very shortly--
  • 00:00:37
    our experts across the space, both from industry,
  • 00:00:41
    non-profits, et cetera.
  • 00:00:44
    So to start with, I want to just really very briefly touch
  • 00:00:47
    on the EUDR.
  • 00:00:48
    I know Remi is going to be talking about this.
  • 00:00:50
    But starting in 2024, companies that
  • 00:00:53
    sell goods, into the European Union,
  • 00:00:55
    are required to prove that these goods were not grown on land
  • 00:00:58
    deforested after December of 2020,
  • 00:01:00
    which is a very large commitment and things--
  • 00:01:04
    and will be difficult for some companies to meet.
  • 00:01:06
    So we're trying to work together to understand how we can help
  • 00:01:11
    companies meet this compliance.
  • 00:01:13
    And these organizations, that will be represented today,
  • 00:01:16
    will be speaking to that.
  • 00:01:18
    As part of the EUDR, there are seven key commodities
  • 00:01:22
    that are--
  • 00:01:22
    and their derivatives that are part of the legislation.
  • 00:01:25
    You can see them up here.
  • 00:01:27
    And the main industries that are impacted
  • 00:01:29
    are, of course, consumer packaged goods,
  • 00:01:32
    retail, furniture, automotive, et cetera.
  • 00:01:36
    So real quickly, today's session,
  • 00:01:38
    we're going to have three parts.
  • 00:01:39
    We're going to hear from our experts.
  • 00:01:41
    They'll have five minutes, each, to speak
  • 00:01:42
    about their perspective.
  • 00:01:45
    We're going to have a quick panel discussion.
  • 00:01:47
    And then open up questions to the audience.
  • 00:01:50
    So with that, let's get started.
  • 00:01:52
    Remi, would you like to come on up.
  • 00:01:55
    REMI D'ANNUNZIO: Thank you.
  • 00:01:56
    Thank you, Alicia.
  • 00:01:57
    Hi, everybody.
  • 00:01:59
    So I'm going to give a really brief introduction to the EU
  • 00:02:06
    Regulation Deforestation and the role
  • 00:02:08
    that FAO is playing in it, somehow.
  • 00:02:12
    I think my question was, what's known about the process
  • 00:02:14
    to be compliant today and what's the role
  • 00:02:16
    of Earth observation and data?
  • 00:02:20
    So I try to put-- so first of all,
  • 00:02:23
    the text of the regulation, itself,
  • 00:02:25
    is fully public, fully available.
  • 00:02:27
    And I encourage you to read it, because it's actually--
  • 00:02:29
    it's long.
  • 00:02:30
    But it's extremely detailed and extremely information-full.
  • 00:02:34
    And there's also, associated to it, a regular,
  • 00:02:37
    Frequently Asked Questions section,
  • 00:02:39
    that is published by the European Commission, that
  • 00:02:41
    is making it easier to read.
  • 00:02:43
    And it's only 20 pages.
  • 00:02:45
    So it really makes it simpler.
  • 00:02:47
    So probably every-- if I have any answers
  • 00:02:50
    to the questions that are asked today,
  • 00:02:51
    I picked them up, already, in the FAQ.
  • 00:02:53
    So I encourage you to read it.
  • 00:02:56
    So I wanted to put some main compliance points.
  • 00:03:00
    The first is the responsibility, on the regulation
  • 00:03:04
    and compliance, is fully on the operators.
  • 00:03:07
    It's not on the countries.
  • 00:03:08
    It's not on the governments.
  • 00:03:09
    It's on the operators that are proposing products
  • 00:03:12
    into the market.
  • 00:03:13
  • 00:03:17
    Key things-- the deforestation, for the regulation,
  • 00:03:24
    means conversions to agricultural land use.
  • 00:03:27
    And degradation means conversions,
  • 00:03:29
    of primary or naturally regenerated forests,
  • 00:03:32
    into plantations.
  • 00:03:33
    And that's that.
  • 00:03:34
    And so you've seen all the agricultural commodities,
  • 00:03:37
    they should not be deforestation--
  • 00:03:40
    with any deforestation.
  • 00:03:41
    And timber should not be with degradation.
  • 00:03:46
    There is something around the geolocation.
  • 00:03:48
    So what is compliant is that all the operators must submit
  • 00:03:53
    the geolocation information, from each farm,
  • 00:03:57
    from where any of the products that enter the market
  • 00:03:59
    come from.
  • 00:04:00
    That's relatively straightforward.
  • 00:04:02
    One thing that is also very important
  • 00:04:04
    is that it should be deforestation-free and legal
  • 00:04:07
    as per the legislation of the country where it comes from.
  • 00:04:10
    And this has not been tackled a lot yet.
  • 00:04:13
    And the cutoff date, we talked about that.
  • 00:04:16
    What's important to remember is that there is no commodity ban.
  • 00:04:20
    There's no country ban, whatever the benchmarking
  • 00:04:22
    system will be.
  • 00:04:25
    There is a benchmarking system, which
  • 00:04:29
    will be interesting to follow in the coming months,
  • 00:04:33
    because it's not published yet.
  • 00:04:34
    So we don't know, yet, what the list of countries
  • 00:04:36
    that are considered standard, low, or high risk.
  • 00:04:39
    But those countries that will be considered standard, low,
  • 00:04:42
    or high risk will have, from the importers,
  • 00:04:45
    from the European countries that get the compliance
  • 00:04:48
    information to them--
  • 00:04:51
    the products that come from standard countries
  • 00:04:53
    will be monitored 3% of the compliance information.
  • 00:04:59
    And the low risk will be lowered down to 1%.
  • 00:05:02
    And high risk will be higher, to 9%.
  • 00:05:04
    So it doesn't mean anything on the operators.
  • 00:05:07
    It means something on the amount of data
  • 00:05:09
    that will be checked by the importing countries.
  • 00:05:12
    And it's important.
  • 00:05:13
    And the European Commission has been trying
  • 00:05:15
    to push the message on this.
  • 00:05:17
    I don't know how long five minutes is.
  • 00:05:19
    But--
  • 00:05:20
    [LAUGHTER]
  • 00:05:23
    --I have like another two minutes?
  • 00:05:25
    ALICIA SULLIVAN: Yes.
  • 00:05:25
    REMI D'ANNUNZIO: OK.
  • 00:05:26
    I just wanted to really briefly talk about that.
  • 00:05:28
    So I'm working in FAO, in the forestry division
  • 00:05:31
    I'm technical lead on the Forest Data Partnership, which
  • 00:05:35
    is funded by USAID and the US State Department,
  • 00:05:38
    with WRI leading, FAO co-leading.
  • 00:05:40
    And we have the immense chance to be partnering,
  • 00:05:44
    with NASA surveyor inside, with Google and, in particular,
  • 00:05:48
    with the Earth Engine team, and with Unilever,
  • 00:05:51
    as a first partner.
  • 00:05:53
    And within the Forest Data Partnership,
  • 00:05:56
    we've been in contact with a dialogue platform called
  • 00:06:01
    DIASCA, that is led by a German corporation and now
  • 00:06:07
    the European Commission, talking about forest monitoring
  • 00:06:10
    and traceability and legality schemes and solutions.
  • 00:06:13
    And we have just received a green light
  • 00:06:15
    to go ahead with the proposal for a digital public
  • 00:06:19
    infrastructure prototype.
  • 00:06:21
    That will be done in collaboration
  • 00:06:23
    with AgStack, who have developed a asset registry.
  • 00:06:26
    I can talk about what the asset registry means.
  • 00:06:28
    But it's basically it's an index system based on the S2 index
  • 00:06:32
    system from Google.
  • 00:06:33
    And that provides basically unique GeoIDs.
  • 00:06:36
    And they've asked us to basically construct
  • 00:06:41
    that line of infrastructure to provide information,
  • 00:06:46
    minimal information, that can be public
  • 00:06:48
    and that can be completely open for anybody.
  • 00:06:50
    And that will probably mean either third-party verification
  • 00:06:52
    or importing countries to have tools,
  • 00:06:54
    at hand, to verify the information that's
  • 00:06:56
    coming in the line.
  • 00:06:59
    I'll be happy to develop furthermore,
  • 00:07:00
    but I just wanted to tell you that this
  • 00:07:03
    is something that's ongoing.
  • 00:07:04
    And more important than the money associated to it,
  • 00:07:08
    what's important is that we've got the mandate, from the EC,
  • 00:07:11
    to develop this, which is sometimes
  • 00:07:13
    extremely tricky to have.
  • 00:07:14
    So we're really happy to move this forward.
  • 00:07:16
    And yeah, we're going to have a collaboration with
  • 00:07:19
    [? Ground, ?] here, from Google, and with the team
  • 00:07:22
    of [? Gino. ?] The AgStack solution, of course,
  • 00:07:24
    is going to be powered by Earth Engine, somehow.
  • 00:07:26
    And this is going to come with the mandate of DIASCA.
  • 00:07:29
    So I'll finish here.
  • 00:07:30
    And I'm really happy to go forward with the questions.
  • 00:07:33
    [APPLAUSE]
  • 00:07:38
    ALICIA SULLIVAN: All right.
  • 00:07:40
    Tasso.
  • 00:07:41
    TASSO AZEVEDO: Good.
  • 00:07:42
  • 00:07:45
    Four things-- these are the four things that we think
  • 00:07:50
    will be key for us to contribute to this discussion using
  • 00:07:58
    the tools that we are preparing.
  • 00:08:00
    So first one is defining what's the scope of application.
  • 00:08:04
    Because it refers to forests, basically, at this point,
  • 00:08:08
    and so you have to check out where
  • 00:08:10
    this will be really applied.
  • 00:08:12
    The second is this whole thing about classifying the risk
  • 00:08:16
    of the country or region.
  • 00:08:17
    So we have kind of an idea of how this could be done,
  • 00:08:23
    not just classifying the whole country
  • 00:08:25
    but thinking on something different.
  • 00:08:27
    Then, obviously, the land use change assessment post 2020.
  • 00:08:32
    And one or two ideas for the traceability,
  • 00:08:35
    which is not exactly what we are working, right now,
  • 00:08:39
    but it can be used.
  • 00:08:40
    Good.
  • 00:08:41
    So for the applications?
  • 00:08:45
    We did a quick view on this, which
  • 00:08:48
    is you need to know where is the forest,
  • 00:08:53
    right, according to the definition,
  • 00:08:55
    in this case, the definition of FAO.
  • 00:08:57
    I think later on, we will have non-wood--
  • 00:09:01
    no, other woodlands, right?
  • 00:09:03
    And then we still are missing the no wood environments,
  • 00:09:08
    right?
  • 00:09:09
    And so why this is important?
  • 00:09:10
    Because, if you see this map, like in the middle,
  • 00:09:14
    you have everything that was forest in 2020.
  • 00:09:18
    In Brazil, we have the classification
  • 00:09:21
    that would be equivalent to the FAO classification of forest.
  • 00:09:24
    And the other one is everything else
  • 00:09:26
    that is not considered forest but is native vegetation also.
  • 00:09:30
    A lot of the Cerrado, for example,
  • 00:09:32
    will be out of this definition.
  • 00:09:34
    So 30% of the deforestation happens on that area.
  • 00:09:41
    But 50% of the conversion for cropland, for example,
  • 00:09:45
    is happening there.
  • 00:09:47
    So basically, by this definition,
  • 00:09:49
    50% of the deforestation or, as you
  • 00:09:53
    say, loss of natural habitats for crops will be simply out.
  • 00:10:01
    And most of the soybean, for example,
  • 00:10:03
    would be out, because they are basically
  • 00:10:04
    being converted in this area.
  • 00:10:06
    So that's why it's very important
  • 00:10:07
    to have this good definition of the scope.
  • 00:10:12
    Second is the classification of risk.
  • 00:10:15
    So there is two things that can be done.
  • 00:10:18
    One is using the historical land cover, land use maps.
  • 00:10:22
    And we have MapBiomas to track what happens
  • 00:10:26
    on the last 10, 15 years or so.
  • 00:10:28
    So you understand how much deforestation
  • 00:10:31
    you have on those areas.
  • 00:10:34
    So for example, in this case here, it's
  • 00:10:37
    the crop land, in general, like in the Cerrado.
  • 00:10:41
    So what you see there is that 35% of the increase
  • 00:10:45
    in cropland, between 2010 and 2020, in the Cerrado,
  • 00:10:49
    was actually on areas that were natural habitats before.
  • 00:10:55
    And forest is the two parts up there.
  • 00:11:00
    So this is a way to have an idea of how much deforestation
  • 00:11:04
    you have on those areas.
  • 00:11:06
    The other thing is that we have these Annual Deforestation
  • 00:11:09
    Reports, which access the potential illegality
  • 00:11:13
    of every single deforestation in Brazil on the last four years.
  • 00:11:16
    So you can have an idea of how much commitments, with rules,
  • 00:11:22
    you have on those areas.
  • 00:11:24
    So for example, if you have an area where people are really
  • 00:11:26
    following the law, if you say that you can't do something,
  • 00:11:30
    it will not be done, right?
  • 00:11:31
    That's more or less the idea.
  • 00:11:33
    But if you are in a region that everybody
  • 00:11:35
    is playing with illegality, you don't
  • 00:11:38
    expect that they will just follow
  • 00:11:40
    the rules because somebody said that they have to do it, right?
  • 00:11:42
    So that's a good way to assess risk.
  • 00:11:44
    And so I think that the way we should translate this
  • 00:11:47
    is that you take a country, like Brazil, which is very big.
  • 00:11:49
    It's not something that you say, look,
  • 00:11:51
    Brazil is all in one place or another place.
  • 00:11:53
    Actually, this varies a lot by crop and by region.
  • 00:11:57
    So the way we see this is that this should evolve
  • 00:12:01
    to something like crops products or commodities versus regions,
  • 00:12:07
    for example.
  • 00:12:08
    So if you take the Amazon, palm oil, cocoa, coffee, and rubber,
  • 00:12:13
    it's definitely very low risk--
  • 00:12:17
    definitely low risk.
  • 00:12:19
    But wood, soy, and cattle will be very high risk.
  • 00:12:23
    And if you go to the Cerrado, wood is not high risk.
  • 00:12:26
    It would be like standard and so on.
  • 00:12:29
    So the idea is that we should think about that a little bit
  • 00:12:33
    more--
  • 00:12:34
  • 00:12:38
    with a little more detail than just, say,
  • 00:12:41
    one country is one place and the other place.
  • 00:12:44
    This is very important, for the incentive, for the people
  • 00:12:47
    to do the right thing.
  • 00:12:48
    Because if you put everybody together, in one spot,
  • 00:12:51
    you make everyone complain.
  • 00:12:53
    And if you say that these are different,
  • 00:12:55
    you have those guys, that are kind of on the blue,
  • 00:12:58
    they will be cheering, for the whole room, on the other side.
  • 00:13:02
    So that's very important for the implementation.
  • 00:13:04
    Yeah.
  • 00:13:05
    This is for the land cover.
  • 00:13:07
    So using the maps, the land cover,
  • 00:13:10
    land use maps that we have, you can identify all change,
  • 00:13:18
    for any direction and for the different types of crops
  • 00:13:21
    or commodities, over the time, for each property, if you want.
  • 00:13:27
    So the deforestation specifically--
  • 00:13:30
    I'm not talking about degradation
  • 00:13:32
    but deforestation, specifically, since we have reports
  • 00:13:35
    on every single deforestation, a report like this
  • 00:13:38
    for every single deforestation that
  • 00:13:40
    happens in Brazil, since January 2019,
  • 00:13:44
    you can locate every deforestation
  • 00:13:46
    on the properties.
  • 00:13:47
    So basically, if you have a property,
  • 00:13:50
    you would know if there is a deforestation there.
  • 00:13:52
    And there is this question whether the application
  • 00:13:55
    of the law is related to the area that you actually claim
  • 00:13:58
    that you plant or the property.
  • 00:14:00
    I must say the only way to make this
  • 00:14:02
    happen is if you consider the property.
  • 00:14:04
    That's the only way you can manage this properly.
  • 00:14:08
    So if you have a deforestation, in the property, after 2020,
  • 00:14:13
    sorry, you don't qualify for [INAUDIBLE]
  • 00:14:16
    because it's too difficult to go beyond that
  • 00:14:18
    in terms of control.
  • 00:14:20
    Yeah.
  • 00:14:20
    And so this is the report, how it does.
  • 00:14:22
    And lastly, for the traceability,
  • 00:14:27
    there is many things on traceability there.
  • 00:14:29
    But, here, we're focusing on things
  • 00:14:31
    that will help you, in large scale,
  • 00:14:34
    to understand if the things make sense or not
  • 00:14:36
    in terms of where the products come from.
  • 00:14:42
    So the first thing is to estimate
  • 00:14:43
    the potential production, of that commodity,
  • 00:14:46
    in a specific property.
  • 00:14:50
    So you check.
  • 00:14:50
    You see the map.
  • 00:14:51
    How much is the cropland that you have there?
  • 00:14:53
    And then you can estimate, oh, it's between this and that,
  • 00:14:56
    that you could have.
  • 00:14:56
    So nobody can claim that they are selling, I don't know,
  • 00:14:59
    five times what you could on that property.
  • 00:15:01
    Because that's a legal property comparing to others,
  • 00:15:04
    for example.
  • 00:15:05
    And the other one is the viability
  • 00:15:07
    of the product to get to a specific processing plant.
  • 00:15:10
    So it's basically, say, you are here, and then
  • 00:15:13
    the guy is saying that he is buying this property,
  • 00:15:15
    from 500 kilometers away.
  • 00:15:17
    This product?
  • 00:15:18
    Forget it.
  • 00:15:19
    That's totally unviable to bring this product
  • 00:15:21
    to the other place.
  • 00:15:22
    So you can kind of--
  • 00:15:25
    And this is not just a radial, right?
  • 00:15:27
    This is really the path.
  • 00:15:29
    So it's finding paths that will give you mileage.
  • 00:15:32
    And this is varied for infrastructure, distance,
  • 00:15:34
    weather, and so on.
  • 00:15:35
    So these are four ways in which we
  • 00:15:38
    can contribute with geo, maps, and et cetera for this thing.
  • 00:15:41
    And let's make it happen, because it makes a difference
  • 00:15:44
    in terms of deforestation.
  • 00:15:47
    [APPLAUSE]
  • 00:15:50
    ALICIA SULLIVAN: All right.
  • 00:15:51
    Now, Pierrick.
  • 00:15:53
  • 00:16:00
    PIERRICK RAMBAUD: Is it working?
  • 00:16:01
    Yeah.
  • 00:16:01
    Perfect.
  • 00:16:02
    I would like to keep Tasso's energy,
  • 00:16:03
    but, with this microphone, I will still
  • 00:16:06
    be in contact with the sound.
  • 00:16:08
    Pierrick Rambaud.
  • 00:16:09
    So I'm working for LDC.
  • 00:16:11
    I know that the crowd is mostly scientists in this conference.
  • 00:16:13
    So before I start, who knows Louis Dreyfus Company
  • 00:16:16
    in the room?
  • 00:16:18
    OK, thank you very much.
  • 00:16:19
    For all the people that doesn't know it,
  • 00:16:20
    it makes my two first slides useful.
  • 00:16:24
    So LDC is a French merchant, on agricultural goods,
  • 00:16:30
    that started working in 1851, so long before I was born.
  • 00:16:36
    We employ around 17,000 employees on 10 platforms.
  • 00:16:40
    I will explain what platforms are, afterward.
  • 00:16:43
    We shipped 80 million tons of commodity,
  • 00:16:47
    on average, annually for a total net sales of $59.9 billion US--
  • 00:16:53
    that's a lot of money--
  • 00:16:54
    in more than 100 countries, over 6 geographic regions.
  • 00:16:59
    So if there's one thing you need to remember,
  • 00:17:01
    we are in the agricultural game for a long time.
  • 00:17:04
    We sell lots of things everywhere.
  • 00:17:08
    So platforms?
  • 00:17:09
    So we have different platforms to actually structure
  • 00:17:11
    the company.
  • 00:17:12
    So some of them are actually commodities-- coffee,
  • 00:17:14
    cotton, food and feed solutions, grains and oilseed, juice,
  • 00:17:19
    rice, and sugar.
  • 00:17:20
    And if you listen carefully to the other speakers,
  • 00:17:22
    you realize that coffee and grains and oilseeds are
  • 00:17:26
    actually-- focusing on EUDR--
  • 00:17:29
    directly challenged by the regulation.
  • 00:17:31
    We have transversal platforms, carbon solutions,
  • 00:17:34
    and global markets, and freight, which
  • 00:17:37
    is dealing with all the vessels that
  • 00:17:39
    are shipping the commodities all over the world.
  • 00:17:44
    This is, more or less, the workflow
  • 00:17:46
    that we need to actually be compliant with EUDR.
  • 00:17:49
    So if anyone, after my presentation,
  • 00:17:52
    have a solution that is, at scale, at countries,
  • 00:17:54
    for any countries and any commodities, please,
  • 00:17:57
    come with us.
  • 00:17:58
    We want you in our team.
  • 00:17:59
    [LAUGHTER]
  • 00:18:01
    So the first step is actually mapping your supply chain.
  • 00:18:05
    That's the first real issue.
  • 00:18:07
    Because mapping your supply chain
  • 00:18:08
    means that you are actually aggregating
  • 00:18:10
    national and partner databases.
  • 00:18:12
    Everything is inconsistent.
  • 00:18:14
    Everything is set differently, so you
  • 00:18:17
    need to align all the metadata, make sure
  • 00:18:19
    that it's actually describing what you want to see.
  • 00:18:22
    And then you need to understand things that are actually
  • 00:18:24
    not dealt by your company.
  • 00:18:27
    So sometimes we bought things from someone else.
  • 00:18:29
    And we need to understand his supply chain.
  • 00:18:33
    Once you get your supply chain, you need to gather the plots.
  • 00:18:36
    Because you know where the commodity is actually
  • 00:18:38
    coming from.
  • 00:18:39
    But you need to define the geometry.
  • 00:18:41
    So as Remi was explaining, less than 4
  • 00:18:42
    hectares, you need one point.
  • 00:18:44
    More than 4 hectares, you need a polygon.
  • 00:18:47
    So then a new nightmare starts.
  • 00:18:49
    You're gathering, again, multiple inconsistent sources.
  • 00:18:52
    What do we have?
  • 00:18:53
    We have public database, cars and buses.
  • 00:18:55
    We have geo-surveys.
  • 00:18:57
    So we send people, on the field, to actually make
  • 00:19:00
    the delineation of our geometries.
  • 00:19:02
    And we have private providers that are just
  • 00:19:04
    doing the same jobs for us.
  • 00:19:06
    Once you get these geometries, you need to check them.
  • 00:19:09
    Because you don't know.
  • 00:19:11
    Are they the truth?
  • 00:19:12
    Are they at the right place?
  • 00:19:13
    Are they underwater bodies?
  • 00:19:15
    You don't know at start.
  • 00:19:17
    So first, you need to make these geometry assessments,
  • 00:19:20
    making sure that they are actually describing the plot.
  • 00:19:23
    And then you also need to check if the geometry in, itself, is
  • 00:19:29
    making sense.
  • 00:19:29
    So anyone that has actually dealt with geometry
  • 00:19:32
    knows that a geometry can be broken.
  • 00:19:35
    Like you track, with a GPS, the coordinates of your field.
  • 00:19:39
    And for some reasons, the GPS just go backward from one point
  • 00:19:43
    to another.
  • 00:19:43
    And boom, your geometry is broken.
  • 00:19:45
    And you cannot perform any analysis anymore.
  • 00:19:48
    So all these checks need to be performed.
  • 00:19:50
    You have your geometries.
  • 00:19:52
    And now you start doing real Earth observation.
  • 00:19:54
    So you need to identify what type of satellite imagery
  • 00:19:59
    are actually useful for this analysis.
  • 00:20:02
    So step one is to leverage public data sets,
  • 00:20:05
    so that we can perform risk assessments, to know which
  • 00:20:08
    countries or which administrative areas are
  • 00:20:10
    actually higher risk.
  • 00:20:11
    And this is where we will do the due diligence.
  • 00:20:14
    And for that, it's still unclear what level of precision
  • 00:20:18
    we need to reach.
  • 00:20:19
    So we don't know if public data sets are sufficient.
  • 00:20:21
    We don't know if we need to use some very high resolution data
  • 00:20:24
    even for the risk assessment.
  • 00:20:26
    So that's an open question.
  • 00:20:27
    And the FAQ of the regulation will reopen soon.
  • 00:20:30
    And that's the type of question that we will ask again.
  • 00:20:34
    We also have questions about the revisiting period,
  • 00:20:37
    because, if you have some commodities that
  • 00:20:39
    are changing, from one year to another,
  • 00:20:40
    you also need to remap your polygon.
  • 00:20:43
    So you need to redo the analysis all over again.
  • 00:20:47
    Identify satellite imagery-- you have your imagery.
  • 00:20:50
    You have your plot.
  • 00:20:51
    And now you're starting to do actual image analysis.
  • 00:20:54
    So you have a field that has been analyzed in your risk
  • 00:20:58
    assessment workflow.
  • 00:21:00
    And it's flagged red.
  • 00:21:02
    There is deforestation according to the forest
  • 00:21:04
    data set, that you are using, for the cutting date.
  • 00:21:08
    We don't want to discard all the fields that
  • 00:21:11
    are set to red on a 10 meter resolution data set.
  • 00:21:14
    Because if we do so, we will lose countless smallholders
  • 00:21:18
    that will be flagged as deforested.
  • 00:21:20
    So there we need to use very high resolution data,
  • 00:21:23
    to perform a real change detection algorithm,
  • 00:21:26
    to really understand if something
  • 00:21:27
    happened on this field.
  • 00:21:28
    And the problem with this very step
  • 00:21:31
    is that the data, that we are using,
  • 00:21:35
    are actually flagging, more or less, half of the fields
  • 00:21:37
    that we are actually analyzing.
  • 00:21:39
    So it means that half of the fields in the world
  • 00:21:41
    need to be mapped with very high resolution data.
  • 00:21:43
    And we need to perform this change detection algorithm
  • 00:21:47
    on half of the fields in the world.
  • 00:21:49
    So our problem, now, is really a commercial problem.
  • 00:21:52
    Because the pricing of doing such an analysis
  • 00:21:55
    is simply prohibitive, even for us, LDC,
  • 00:21:58
    that is one of the big players in the agricultural market.
  • 00:22:01
    So that's one huge issue that is not yet completely answered.
  • 00:22:06
    And if you done everything, you know what field is actually
  • 00:22:09
    deforested, what field is not deforested,
  • 00:22:12
    you still have one step to actually manage,
  • 00:22:15
    which is taking action.
  • 00:22:16
    Because, OK, we know that this field is deforested.
  • 00:22:18
    But what is going to happen to this farmer?
  • 00:22:20
    So we have different options.
  • 00:22:22
    The first one, of course, is just discard this farmer
  • 00:22:26
    from the EU market, which leads to a segregation issue.
  • 00:22:31
    When the goods are actually coming to the mills
  • 00:22:33
    or to the crushing plants, you need
  • 00:22:35
    to segregate it from the rest, so
  • 00:22:36
    having a pile of bags of coffee that
  • 00:22:39
    are EU compliant and another one that is EU non-compliant.
  • 00:22:42
    And that's one of the key issue that we are
  • 00:22:44
    facing within our supply chain.
  • 00:22:47
    And we are still asking the European Commission
  • 00:22:50
    if there will be any mitigation solutions so that farmers that
  • 00:22:54
    have been removed from the European market
  • 00:22:57
    could re-enter this market by performing some extra work
  • 00:23:02
    replanting, regrowing on their deforested fields.
  • 00:23:06
    So again, anyone that has this, please come to me.
  • 00:23:09
    I will be super happy to discuss.
  • 00:23:11
  • 00:23:14
    [APPLAUSE]
  • 00:23:17
    BOBBY PINTER: Yeah.
  • 00:23:18
    Thank you.
  • 00:23:19
    So I'm not going to pretend we have that,
  • 00:23:21
    but that is the direction I think we are going is NGIS.
  • 00:23:24
    So hopefully, I'll take everything
  • 00:23:26
    you've seen, just now, and provide it
  • 00:23:27
    in the context of a story on how we're
  • 00:23:29
    trying to work with our customers and partners.
  • 00:23:31
    So NGIS is an organization that's
  • 00:23:33
    been partnered with Google for over 10 years.
  • 00:23:35
    NGIS is an Australian geospatial consultancy.
  • 00:23:38
    A lot of the work that we do, really,
  • 00:23:40
    in partnership with Google, is focusing
  • 00:23:42
    on making sure that Earth Engine and other geospatial tools
  • 00:23:44
    are reaching customers and especially our joint customers.
  • 00:23:47
    So there's really three areas in which
  • 00:23:49
    we're partnered with Google.
  • 00:23:50
    I'll go through these, fairly briefly,
  • 00:23:51
    because we've got a lot to cover on the EUDR side.
  • 00:23:53
    So we really think about solutions
  • 00:23:55
    co-development and go-to-market.
  • 00:23:56
    How do we take Earth Engine and some of these core technologies
  • 00:23:58
    and bring them into the kinds of systems and workflows
  • 00:24:01
    that organizations can use?
  • 00:24:02
    We're also a Cloud Build, Sell & Service Partner.
  • 00:24:04
    Not necessarily relevant to all of you in the room,
  • 00:24:06
    but the idea is that we know what
  • 00:24:07
    we're doing with Google Cloud, and we like to use as much
  • 00:24:09
    of Google Cloud, as possible, in our solutions.
  • 00:24:11
    And finally, we really think about ourselves
  • 00:24:13
    as being aligned to Google's one gigaton goal.
  • 00:24:15
    So we invest in some of the programs,
  • 00:24:17
    like the Google Earth Engine Publisher
  • 00:24:18
    Program, which you may have seen me present on,
  • 00:24:20
    earlier, yesterday, and just other of these kinds
  • 00:24:23
    of initiatives and really how we're partnering with Google.
  • 00:24:25
    So a little bit about us and the things that we build.
  • 00:24:28
    I'll be spending all of my time, really,
  • 00:24:29
    today talking about Tracemark.
  • 00:24:30
    And the idea is, really, in partnership with Google--
  • 00:24:33
    and, actually, we are the implementation partner
  • 00:24:35
    for some of the partnership work between Google and Unilever.
  • 00:24:38
    Some of the work, that we're trying to accomplish,
  • 00:24:39
    is around how do we actually build out
  • 00:24:41
    those kinds of systems that were just
  • 00:24:43
    explained in so much detail for these types of organizations.
  • 00:24:45
    But I'm going to spend my time really
  • 00:24:47
    talking about the complexities that organizations might face.
  • 00:24:49
    And I'm going to use coffee as my main example.
  • 00:24:51
    So a little bit about coffee.
  • 00:24:53
    This is a space where I work in, quite frequently.
  • 00:24:56
    So this is kind of close to my heart.
  • 00:24:57
    But coffee is normally exported from a country of origin.
  • 00:25:00
    It's normally processed.
  • 00:25:01
    It's an unroasted bean.
  • 00:25:02
    For those of you who are coffee nerds,
  • 00:25:04
    this is often called a green bean or something else
  • 00:25:05
    like that.
  • 00:25:06
    So a coffee purchaser will purchase
  • 00:25:08
    a number of different beans produced by farmers.
  • 00:25:10
    And they're thinking about things like taste, quality,
  • 00:25:12
    and varietal.
  • 00:25:12
  • 00:25:15
    Sorry, I skipped a slide there for a sec.
  • 00:25:17
    So when you think about coffee, there's
  • 00:25:19
    things like single origin, which you may have heard about.
  • 00:25:21
    And there's other things like coffee blends.
  • 00:25:22
    And I've made up this entire diagram,
  • 00:25:24
    so don't assume this means anything in particular.
  • 00:25:26
    But the idea is it's really comprised
  • 00:25:28
    of a number of different beans that
  • 00:25:29
    are trying to create a consistency and a flavor
  • 00:25:31
    profile.
  • 00:25:31
    If you're buying a bag of beans from the store,
  • 00:25:33
    you want it to taste roughly the same every time.
  • 00:25:35
    However, that's a challenge.
  • 00:25:36
    We have smallholder farmers.
  • 00:25:38
    There's hundreds of thousands of various coffee farmers, who
  • 00:25:41
    might be in an overall ecosystem,
  • 00:25:42
    so each lot, even if it's the same SKU,
  • 00:25:45
    is going to have beans from different locations.
  • 00:25:48
    And then roasting is normally done in a country of origin.
  • 00:25:50
    So there's this massive aggregation problem.
  • 00:25:52
    So if a company is buying a large number of beans
  • 00:25:55
    and they're thinking about, OK, we've got a silo
  • 00:25:57
    or we've got some other way of containing these beans that's
  • 00:26:00
    thousands of pounds, that might be from thousands of locations.
  • 00:26:03
    And when we think about that in the context EU deforestation,
  • 00:26:05
    this really represents a sizable challenge.
  • 00:26:08
    So really, the two parts of the deforestation regulation--
  • 00:26:10
    this has been covered, already, a little bit now--
  • 00:26:12
    is there's these two requirements that we're focused
  • 00:26:15
    on--
  • 00:26:15
    the geolocation of the plots of land and then
  • 00:26:17
    also ensuring, that there's conclusive
  • 00:26:18
    and verifiable information, that the products are
  • 00:26:21
    deforestation-free.
  • 00:26:21
    You've heard a lot about that, so no need
  • 00:26:23
    to elaborate further.
  • 00:26:24
    But in our opinion, this really highlights
  • 00:26:26
    the crux of the problem, especially for an organization.
  • 00:26:28
    They have to perform due diligence
  • 00:26:30
    by combining their transactional data,
  • 00:26:32
    around what beans are roasted as part of a given lot,
  • 00:26:35
    where and when are those beans sourced within the supply
  • 00:26:37
    chain, and then, also, those sustainability
  • 00:26:40
    facts and insights around deforestation risk
  • 00:26:42
    and other risks and social risks,
  • 00:26:44
    really, around the ability to produce beans there.
  • 00:26:47
    So I think this really highlights
  • 00:26:48
    what we see as the problem.
  • 00:26:49
    If your due diligence statement, for example,
  • 00:26:51
    covers 10,000 pounds of beans, and it turns out
  • 00:26:55
    that one of those sources is no good,
  • 00:26:58
    many of these organizations are saying,
  • 00:26:59
    we may have to just throw out the entire 10,000 pounds.
  • 00:27:02
    Because the way that the regulation is written
  • 00:27:04
    is, unless you can properly keep that verified,
  • 00:27:07
    we're talking about either massive loss,
  • 00:27:09
    recall in some scenarios-- very costly scenarios
  • 00:27:12
    that organizations are looking at to say,
  • 00:27:13
    this is quite serious how business will be run,
  • 00:27:16
    the risk of having to go through these very expensive
  • 00:27:19
    operational challenges.
  • 00:27:20
    Entirely unrelated to the science, itself, just how would
  • 00:27:23
    an organization pull back 10,000 pounds of beans
  • 00:27:26
    or throw out entire shipping containers worth of beans.
  • 00:27:29
    This is the scope and the magnitude
  • 00:27:31
    of the business challenge.
  • 00:27:32
    So we see our opportunity as saying,
  • 00:27:34
    take this amazing scientific community and the other folks,
  • 00:27:36
    who are, here, working on what is the definition of forest,
  • 00:27:39
    how do we know what deforestation is and isn't, and
  • 00:27:42
    bring some of that complexity in a way that's
  • 00:27:44
    a little bit simpler for these organizations
  • 00:27:46
    to bring together.
  • 00:27:47
    So helping them focus on how do we
  • 00:27:49
    take our transactional data, our view of the problem,
  • 00:27:52
    and then marry that with the science and the sustainability
  • 00:27:54
    insights that we're bringing together.
  • 00:27:56
    So some of the areas that we're focused on, today.
  • 00:27:58
    As mentioned, we're working with Unilever.
  • 00:28:00
    So we're doing a lot in the palm oil related space.
  • 00:28:02
    We're in fairly advanced conversations across a number
  • 00:28:05
    of these different commodities.
  • 00:28:06
    This is an evolving problem.
  • 00:28:07
    There is a lot of things that we're bringing today,
  • 00:28:10
    in how we work and partner with organizations
  • 00:28:12
    to create UIs and dashboards and take advantage of that science.
  • 00:28:15
    But we see ourselves, over the next year,
  • 00:28:17
    really making sure that this effort is done, in partnership,
  • 00:28:19
    both on the science side, around making sure we're bringing
  • 00:28:22
    in the right map layers and the data products
  • 00:28:24
    that you've already heard about, for the last 10 or so minutes.
  • 00:28:26
    But then making sure that we work with enterprises
  • 00:28:28
    around-- how will you actually take this data
  • 00:28:30
    and make it useful and hopefully avoid
  • 00:28:31
    some of these really negative business
  • 00:28:33
    outcomes that we've spent some time on already.
  • 00:28:35
    So thank you so much for the time.
  • 00:28:36
    I'm looking forward to more conversation.
  • 00:28:38
    [APPLAUSE]
  • 00:28:43
    EVAN TACHOVSKY: Hey, everybody.
  • 00:28:44
    My name is Evan Tachovsky.
  • 00:28:45
    I'm the global director of the Data Lab at WRI.
  • 00:28:48
    I'm a data scientist by training.
  • 00:28:49
    I'm not a policy person.
  • 00:28:51
    So I'm going to talk to you, today, a little bit about what
  • 00:28:52
    we can do, as a technical community,
  • 00:28:54
    to support all of this important work you've seen here.
  • 00:28:56
    So many of you will actually work
  • 00:28:58
    on projects and organizations and teams that work directly
  • 00:29:00
    on EUDR.
  • 00:29:01
    But many more of you will build technology or train people,
  • 00:29:04
    train students, and work with others who will go on
  • 00:29:07
    to carry on this legacy.
  • 00:29:08
    And so I wanted to talk a little bit
  • 00:29:09
    about what we can do, as a community,
  • 00:29:10
    to actually make sure that we have
  • 00:29:12
    the geospatial infrastructure tools
  • 00:29:15
    and support to actually implement and follow
  • 00:29:16
    through on this.
  • 00:29:17
    So at WRI, of course, we have teams that produce data layers,
  • 00:29:21
    teams that produce platforms, and teams that work directly
  • 00:29:24
    with a number of companies in similar ways
  • 00:29:26
    that others described here.
  • 00:29:28
    But when we think about the next generation of tools,
  • 00:29:30
    there are three things that we're actually
  • 00:29:31
    shifting towards.
  • 00:29:32
    So the first is not just platforms but open
  • 00:29:35
    infrastructure.
  • 00:29:36
    This is key and was highlighted in Remi's.
  • 00:29:38
    The digital public infrastructure
  • 00:29:40
    to enable a whole bunch of other platforms to work
  • 00:29:43
    is more important than just the success of one platform.
  • 00:29:47
    When you get into this, we have, I
  • 00:29:48
    think, a strong view coming from the startup space
  • 00:29:51
    that someone has to win the market.
  • 00:29:52
    When you think about all the problems
  • 00:29:54
    we're highlighting here, there isn't just one market.
  • 00:29:56
    There's a million sort of submarkets
  • 00:29:59
    and problems and areas.
  • 00:30:01
    And so it's going to take a number of different solutions,
  • 00:30:03
    solutions built, here, in Silicon Valley, but also
  • 00:30:06
    solutions built in the global South.
  • 00:30:08
    And our infrastructure, we believe,
  • 00:30:10
    is what's needed to unlock that.
  • 00:30:12
    So very specifically what does that mean?
  • 00:30:13
    That means work today on foundational models
  • 00:30:16
    is incredibly important for geospatial.
  • 00:30:18
    Because those models will be what
  • 00:30:20
    is needed for others, who maybe don't have access
  • 00:30:22
    to the scale of compute or the training,
  • 00:30:24
    to actually build models that work for their communities.
  • 00:30:27
    And so if you're working on a foundational model,
  • 00:30:29
    today, or you're working towards those efforts,
  • 00:30:31
    you're helping us and future us work toward compliance
  • 00:30:33
    with the EUDR.
  • 00:30:34
    The second thing is we believe we're
  • 00:30:37
    going more towards tailored apps and very specific use cases.
  • 00:30:42
    I think in the Geo for Good community
  • 00:30:43
    and the general Data for Good community,
  • 00:30:46
    we've come from a place of big platforms.
  • 00:30:48
    We have one at WRI.
  • 00:30:49
    We have this thing called Resource Watch they're actually
  • 00:30:50
    in the process of winding down and decommissioning.
  • 00:30:53
    500 data sets, a million map layers, gidgets,
  • 00:30:57
    gizmos everywhere, and the usage isn't there.
  • 00:30:59
    I'll tell you honestly, the usage isn't there.
  • 00:31:02
    Where is the usage?
  • 00:31:03
    A very simple platform called GFW Pro,
  • 00:31:06
    which is a version of our GFW that's
  • 00:31:08
    made for the needs, the specific needs of commodity firms.
  • 00:31:11
    And it's extremely simple in the functionality.
  • 00:31:13
    It allows an API interface for folks
  • 00:31:15
    to give us data on points, get back various information
  • 00:31:18
    from GFW.
  • 00:31:19
    And then there are other functionalities.
  • 00:31:20
    But it's very simply focused on this.
  • 00:31:22
    And the uptake we see with that is massive.
  • 00:31:25
    Similarly, on the other end of the spectrum,
  • 00:31:27
    we have a mobile application called Forest Watcher.
  • 00:31:29
    And Forest Watcher is meant to put this information
  • 00:31:31
    in the hands of people who are frontline forest
  • 00:31:33
    defenders, journalists who might be investigating things,
  • 00:31:36
    totally the other end of the spectrum from big consumer good
  • 00:31:38
    companies.
  • 00:31:39
    And that works for them, because it's in the size and shape
  • 00:31:42
    and really tailored to their needs.
  • 00:31:44
    Get some of the same information that folks
  • 00:31:46
    are getting on GFW Pro but in a totally different package,
  • 00:31:49
    totally different usability, and totally different languages,
  • 00:31:52
    and a whole bunch of other things
  • 00:31:53
    that actually make it really accessible.
  • 00:31:54
    The third thing that we think is important,
  • 00:31:56
    that we're shifting more towards,
  • 00:31:57
    is community building and support.
  • 00:32:00
    This law doesn't work and isn't equitable without actually
  • 00:32:03
    building communities of folks, on the ground,
  • 00:32:05
    who are benefiting from this regulation.
  • 00:32:08
    This is in addition to being a massive important experiment
  • 00:32:11
    for climate, this is one of the biggest economic experiments
  • 00:32:14
    in the world.
  • 00:32:15
    We're going into every supply chain, people's livelihoods,
  • 00:32:18
    and changing things, significantly.
  • 00:32:20
    And if we don't empower communities
  • 00:32:22
    to understand this law, to understand their own data,
  • 00:32:24
    to understand the things that we'll be asking them for,
  • 00:32:27
    as a global community of people who
  • 00:32:28
    work for big companies, people who work for big organizations,
  • 00:32:31
    they'll be left out.
  • 00:32:32
    And as the geospatial community, we
  • 00:32:34
    have a profound tradition of making things better.
  • 00:32:36
    But we also have made things worse.
  • 00:32:39
    Way back in grad school, I did a lot of work
  • 00:32:40
    on redlining in the US.
  • 00:32:42
    And you see that applications of risk frameworks and things
  • 00:32:45
    that are meant entirely perfectly in good stead.
  • 00:32:49
    Redlining, here, in the US, was a sort of insurance framework.
  • 00:32:52
    It's very straightforward.
  • 00:32:54
    You need to have insurance.
  • 00:32:55
    And so we need frameworks for that.
  • 00:32:56
    But that led to unintended consequences of generations
  • 00:32:59
    of disenfranchisement.
  • 00:33:01
    And if we don't empower communities,
  • 00:33:02
    we don't train communities and support communities,
  • 00:33:04
    we're at risk of being well-intentioned but ultimately
  • 00:33:07
    ending up with outcomes that I don't think anyone in this room
  • 00:33:09
    would want.
  • 00:33:09
    And so those three things--
  • 00:33:11
    of course, we have the same sort of stuff
  • 00:33:13
    we've always done at WRI.
  • 00:33:14
    We publish reports, build tools.
  • 00:33:16
    But increasingly shifting to public infrastructure,
  • 00:33:19
    tailored apps, very specific things,
  • 00:33:21
    and then building and supporting communities
  • 00:33:22
    are where we're going.
  • 00:33:23
    And I'm so excited to do that with all of you.
  • 00:33:25
    So thanks.
  • 00:33:26
    [APPLAUSE]
  • 00:33:29
    ALICIA SULLIVAN: All right.
  • 00:33:31
    Big thank you for everyone who spoke.
  • 00:33:34
    So now we're going to-- and we go until 2:45, correct?
  • 00:33:39
    OK, good.
  • 00:33:40
    So let me find my notes, here.
  • 00:33:42
  • 00:33:46
    So we're going to have a panel.
  • 00:33:47
    We're going to have a panel discussion.
  • 00:33:49
    Where is my cursor?
  • 00:33:50
  • 00:33:55
    Oh, sorry, my cursor ran away.
  • 00:34:00
    There it is.
  • 00:34:02
    OK.
  • 00:34:04
    Yeah.
  • 00:34:06
    Why don't you guys come on up here.
  • 00:34:07
    And we can have a quick discussion,
  • 00:34:10
    and then we'll go to audience questions.
  • 00:34:13
    And there's another microphone, there, on the table.
  • 00:34:15
    PIERRICK RAMBAUD: Should I bring one?
  • 00:34:17
    ALICIA SULLIVAN: Yeah.
  • 00:34:17
    Or you guys can stand up here, either way.
  • 00:34:19
    [LAUGHS] Awesome.
  • 00:34:21
    Awesome.
  • 00:34:21
    Awesome.
  • 00:34:22
  • 00:34:25
    That's totally fine.
  • 00:34:28
    All right, so there's also another microphone, over there,
  • 00:34:31
    on the table.
  • 00:34:32
    Maybe, Remi, you can grab that.
  • 00:34:35
    Great.
  • 00:34:37
    REMI D'ANNUNZIO: Releasing a Marvel movie.
  • 00:34:39
    ALICIA SULLIVAN: I know.
  • 00:34:39
    I know.
  • 00:34:40
    Thank you for being flexible, here.
  • 00:34:42
    All right, so I have some questions.
  • 00:34:45
    But I'm also happy to, if you guys have other questions
  • 00:34:47
    for each other or things like that, that come up
  • 00:34:50
    in the course of discussion, I'm happy to go off script a bit,
  • 00:34:52
    as well.
  • 00:34:53
    So the first question I have is, where
  • 00:34:55
    do you think the biggest gap is, now, for companies
  • 00:34:58
    that need to comply with this regulation?
  • 00:35:00
    And I'd love for Pierrick and maybe Bobby to start with that.
  • 00:35:03
  • 00:35:06
    PIERRICK RAMBAUD: Is it?
  • 00:35:07
    Yes, it's working.
  • 00:35:09
    Biggest gap within this regulation, for us,
  • 00:35:13
    is, A, the identification of the field geometries.
  • 00:35:19
    That's the most difficult part.
  • 00:35:21
    There are many data sets available with geometries
  • 00:35:24
    that are predefined.
  • 00:35:25
    And usually, they are actually always related
  • 00:35:27
    to farms, farms that actually grow different commodities.
  • 00:35:31
    And the EU regulation is commodity-based.
  • 00:35:33
    So once you have--
  • 00:35:35
    I was talking about CAR previously.
  • 00:35:37
    So CAR is referencing all the farms in Brazil.
  • 00:35:41
    Correct me if I'm wrong.
  • 00:35:43
    And this is farm-related.
  • 00:35:44
    So if you're growing different commodities,
  • 00:35:46
    but you only actually want to get the specific delineation
  • 00:35:50
    for soy, you need to go to this CAR number,
  • 00:35:53
    split it again, manually or automatically,
  • 00:35:56
    and identify each crop that exists within this field.
  • 00:36:00
    And that's great, because Brazil have CAR.
  • 00:36:03
    But there are many countries that have nothing.
  • 00:36:05
    And there you're all alone with Earth observation
  • 00:36:08
    automatic tools trying to find the fields within the forest.
  • 00:36:11
    Because coffee is a shaded commodity.
  • 00:36:14
    And it grows within other trees.
  • 00:36:16
    So it looks exactly like forest.
  • 00:36:18
    So that's the main gap for us.
  • 00:36:22
    BOBBY PINTER: Yeah, I would say it's
  • 00:36:24
    the same exact thing that we think about and we talk about.
  • 00:36:27
    I think the challenge, especially from a larger
  • 00:36:29
    organization perspective, is normally
  • 00:36:31
    there's a number of different tiers in their supply chain.
  • 00:36:34
    And they may have certain relationships
  • 00:36:35
    to a certain point.
  • 00:36:36
    And then they expect that their suppliers have
  • 00:36:38
    to go talk to their suppliers.
  • 00:36:40
    And you get this very convoluted game of telephone.
  • 00:36:42
    And as you were mentioning, it's not necessarily
  • 00:36:45
    a remote sensing solvable problem in all cases.
  • 00:36:48
    Like there isn't going to be some capability that
  • 00:36:50
    allows us to know.
  • 00:36:51
    Just because I can see that it's a farm,
  • 00:36:53
    from space, does not mean I know who they sell to.
  • 00:36:56
    What happens when it goes into a shipping container
  • 00:36:59
    or to some other facility?
  • 00:37:00
    How does it actually reach those points in the supply chain?
  • 00:37:03
    So the linkages is a big problem.
  • 00:37:05
    I think the other part is going to be
  • 00:37:06
    around the balancing of what is seen
  • 00:37:11
    as a little bit of security and IP
  • 00:37:13
    especially around sharing some of these sourcing locations.
  • 00:37:16
    There's a bit of a dynamic shift that
  • 00:37:19
    needs to happen in terms of engagement,
  • 00:37:21
    with farmers and with those who are actually
  • 00:37:23
    maintaining these areas.
  • 00:37:25
    It's very difficult, right now, because you ask questions
  • 00:37:27
    that they either don't understand why they're
  • 00:37:29
    answering or they don't see enough value in answering
  • 00:37:31
    some of the questions.
  • 00:37:32
    Many companies have their own processes and certifications
  • 00:37:35
    and surveys.
  • 00:37:36
    So you're saying, how much effort
  • 00:37:38
    do I have to go through in order to actually follow or check
  • 00:37:42
    all the boxes in someone's various checklists?
  • 00:37:44
    And it's a big administrative burden
  • 00:37:46
    especially for these smaller organizations.
  • 00:37:49
    So there's a bit of a dynamic shift
  • 00:37:50
    that I think we're trying to contend against there.
  • 00:37:53
  • 00:37:55
    TASSO AZEVEDO: Two things--
  • 00:37:56
    I think one thing is, as I said before, I
  • 00:38:00
    don't think this will be solved if we
  • 00:38:02
    try to do parcel by parcel.
  • 00:38:04
    It has to be the property.
  • 00:38:07
    Anything else would be just a guess, because you say, oh,
  • 00:38:11
    you can clear-cut the forest, here,
  • 00:38:12
    for this crop that is not--
  • 00:38:14
    If the objective of the law is to stop deforestation,
  • 00:38:19
    I mean, it's not a question.
  • 00:38:21
    This parcel, here, is for soybean.
  • 00:38:23
    There's no new deforestation.
  • 00:38:25
    But I keep clear-cutting this for other crops.
  • 00:38:28
    It makes no sense.
  • 00:38:31
    I mean you can even take this as a decision.
  • 00:38:33
    Say for your control, you say, well,
  • 00:38:36
    I'm simply considering the property,
  • 00:38:38
    because it's much safer than trying to find out the plot.
  • 00:38:41
    So I think this could be--
  • 00:38:42
    The other thing about the chain of custody,
  • 00:38:44
    I think there is a lot of lessons,
  • 00:38:47
    from the whole process of FSC certification, because we
  • 00:38:51
    spent 30 years, almost, like building up this thing.
  • 00:38:53
    And the end result of that is the only way to do it
  • 00:38:57
    is mass balance.
  • 00:38:59
    We will not find a way to do it, like tracking
  • 00:39:01
    every truck and et cetera.
  • 00:39:04
    And actually, this will not help the objective.
  • 00:39:08
    If our objective is really the deforestation,
  • 00:39:11
    than a simple mass balance control by region
  • 00:39:15
    will help us to get through.
  • 00:39:18
    At least, on the beginning, just make
  • 00:39:20
    something that is simple mass balance.
  • 00:39:23
    And these two things are very important.
  • 00:39:25
    Because, as much as we make this complicated,
  • 00:39:28
    we have an argument to postpone the beginning.
  • 00:39:33
    Because we will arrive at the end of 2024,
  • 00:39:35
    people will say, oh, it's impossible to do it.
  • 00:39:37
    Postpone.
  • 00:39:38
    So we have to do something that is much more simple to start.
  • 00:39:43
    And then, OK, we can start.
  • 00:39:45
    And then we go and kind of a lock-in as we go on with this.
  • 00:39:50
    ALICIA SULLIVAN: That's really interesting.
  • 00:39:52
    And I'm wondering maybe, Remi, would the EU Commission
  • 00:39:57
    be open to taking an approach, like Tasso was describing,
  • 00:40:01
    that is a little bit more of a step towards the way
  • 00:40:05
    that the regulation is written today?
  • 00:40:07
  • 00:40:10
    REMI D'ANNUNZIO: So first thing, maybe the really long term
  • 00:40:16
    objective of the European Commission
  • 00:40:18
    is to eliminate deforestation.
  • 00:40:19
    But at the moment, the only objective
  • 00:40:20
    of the European Commission is to prevent
  • 00:40:22
    products that are deforestationary to enter
  • 00:40:24
    the EU market.
  • 00:40:25
    And they don't really care what's
  • 00:40:26
    going to happen otherwise.
  • 00:40:27
    And to be honest, there's a lot of collateral damage
  • 00:40:30
    that's going to happen on some commodities,
  • 00:40:32
    just because of the definition.
  • 00:40:33
    And shaded crops, let's just be clear, coffee, cocoa--
  • 00:40:38
    because the main aspect of the definition on deforestation
  • 00:40:42
    is the change on land use.
  • 00:40:44
    So you enter forest, if it's classified as a forest.
  • 00:40:48
    You enter it.
  • 00:40:49
    You enter any crop in it, it's considered deforestation.
  • 00:40:52
    And that cannot enter the market anymore.
  • 00:40:54
    That's done.
  • 00:40:54
    So my first advice would be to actually
  • 00:40:57
    look at economists and legal people,
  • 00:41:00
    because there's a lot of those products
  • 00:41:02
    that we have to go to other markets.
  • 00:41:04
    Because, I think, that's going to be
  • 00:41:05
    a reality in December 2024.
  • 00:41:13
    I think there is definitely a stepwise approach that's
  • 00:41:16
    going to happen.
  • 00:41:19
    There's a lot of specifications that are not existing yet.
  • 00:41:23
    So it's still an ongoing process.
  • 00:41:27
    What you mentioned around Cerrado
  • 00:41:29
    is something that is already considered.
  • 00:41:31
    They already consider including other wooded lands
  • 00:41:34
    in the upcoming iterations, next year, of the regulation.
  • 00:41:38
    So they do have that stepwise approach in mind.
  • 00:41:44
    And what can I say?
  • 00:41:46
    No.
  • 00:41:47
    Yeah.
  • 00:41:49
    I think we should be very conscious that, I think,
  • 00:41:52
    the EC wanted to target the big bulk of deforestation
  • 00:41:57
    associated to agricultural development.
  • 00:41:59
    They want to prevent that to enter the market
  • 00:42:02
    and took some decisions to make it relatively clear in terms
  • 00:42:05
    of semantics and definitions.
  • 00:42:08
    And because of that, there will be
  • 00:42:10
    collateral damage on markets.
  • 00:42:12
    And there will be some change in the supply chains.
  • 00:42:15
    ALICIA SULLIVAN: Thank you very much.
  • 00:42:16
    REMI D'ANNUNZIO: I imagine you want this one.
  • 00:42:17
    ALICIA SULLIVAN: Yes, please.
  • 00:42:19
    PIERRICK RAMBAUD: So just to jump on the definition problem,
  • 00:42:24
    there is one issue that, as a private company,
  • 00:42:28
    we cannot actually handle, which is the definition of what was
  • 00:42:31
    forest in 2020.
  • 00:42:33
    There are hundreds of data sets that
  • 00:42:35
    are defining what is forest, what is non-forest.
  • 00:42:38
    Most of them are actually not defining
  • 00:42:40
    what's forest and non-forest but what
  • 00:42:41
    is tree and not-tree-- so back again, palm trees, coffee,
  • 00:42:44
    cocoa.
  • 00:42:45
    But at the moment, the European Commission has not said,
  • 00:42:49
    this data set is the truth.
  • 00:42:50
    And all the others, you can use them,
  • 00:42:52
    but they are just helping you building your forest
  • 00:42:57
    probability mask.
  • 00:42:58
    So as long as we don't really know which is the reference
  • 00:43:02
    data set-- and I'm not sure that they will actually select one--
  • 00:43:06
    every private company will select his provider,
  • 00:43:09
    will select his data set, and send,
  • 00:43:11
    to the EU, due diligence report that will
  • 00:43:14
    be based on different sources.
  • 00:43:16
    And that's an inconsistency that we
  • 00:43:18
    don't know how the European Commission will actually
  • 00:43:20
    handle.
  • 00:43:21
    [LAUGHTER]
  • 00:43:22
    REMI D'ANNUNZIO: Can I just say that I'm pretty sure they
  • 00:43:25
    won't.
  • 00:43:25
    I'm pretty sure they will not take any responsibility
  • 00:43:28
    in saying this is the data set to use.
  • 00:43:30
    You have to know that.
  • 00:43:32
    And yes, you're right.
  • 00:43:34
    Until then and as long as the regulations are going to be on,
  • 00:43:37
    it's going to be your responsibility
  • 00:43:39
    to provide the due diligence.
  • 00:43:40
    And then that's going to be their problem to verify.
  • 00:43:44
    TASSO AZEVEDO: Just to be honest,
  • 00:43:46
    it doesn't really matter.
  • 00:43:47
    ALICIA SULLIVAN: It's a great question.
  • 00:43:49
    TASSO AZEVEDO: Because it's like 80%.
  • 00:43:51
    I mean, if we can tackle 80%, 90% of the problem,
  • 00:43:54
    that's done.
  • 00:43:55
    Right?
  • 00:43:56
    You don't have to be guaranteed that this will be like 100%.
  • 00:43:59
    So if you have your map, your reference, and it's working,
  • 00:44:03
    maybe there will be an error, here or there.
  • 00:44:07
    Just we have to think about the scale of the problem.
  • 00:44:10
    PIERRICK RAMBAUD: That's good for you.
  • 00:44:11
    We are fine now.
  • 00:44:13
    REMI D'ANNUNZIO: No, no.
  • 00:44:14
    No, no, no, what I'm saying is like, if the regulation says
  • 00:44:16
    that your map is fine, whatever.
  • 00:44:18
    The map that you present is fine.
  • 00:44:20
    I mean, what I'm saying is that, for the objective, it's fine.
  • 00:44:24
    It's fine that you have different maps.
  • 00:44:26
    Because the different maps will not be so different.
  • 00:44:28
    They will be different.
  • 00:44:29
    There will be things here and there.
  • 00:44:31
    And for the big thing, think of the big picture.
  • 00:44:34
    Basically, we have 4 million hectares
  • 00:44:37
    of deforestation every year.
  • 00:44:38
    Those commodities are representing 80% of them.
  • 00:44:40
    If we solve 80%, I mean we take down 60% of the deforestation.
  • 00:44:45
    That's an amazing achievement that we can have
  • 00:44:48
    forced by one small market.
  • 00:44:50
    Because comparably, the market is not even that big.
  • 00:44:53
    But companies your size will have
  • 00:44:55
    to apply the same kind of practices whatever
  • 00:44:59
    market you go, because, if you go to the higher tier,
  • 00:45:02
    it will happen.
  • 00:45:03
    So I think this is like a--
  • 00:45:04
    yeah.
  • 00:45:05
    So--
  • 00:45:06
    ALICIA SULLIVAN: I guess this is a somewhat nice segue
  • 00:45:08
    into another question I had for the panel.
  • 00:45:11
    What role do you see commercial data or solutions providers
  • 00:45:14
    in supporting companies to meet this regulation?
  • 00:45:16
    And as a follow-up question, what
  • 00:45:19
    degree of scientific scrutiny and transparency do you think
  • 00:45:22
    will be needed from those commercial providers?
  • 00:45:25
    And maybe, Evan, do you want to start with that?
  • 00:45:27
    EVAN TACHOVSKY: Sure, go to the non-commercial person, too.
  • 00:45:29
    [LAUGHTER]
  • 00:45:30
    So we have a saying at WRI, when do know when to stop.
  • 00:45:34
    And we stop when the major consultancies come in
  • 00:45:37
    as a general rule.
  • 00:45:38
    And so if a major consulting firm
  • 00:45:40
    feels there's enough market, enough standardization, enough
  • 00:45:42
    kind of thing there, you probably
  • 00:45:44
    don't actually need us.
  • 00:45:45
    Now, in this case, I think that's partially true
  • 00:45:47
    but partially not true in that I think
  • 00:45:49
    there will be a lot of providers who step up
  • 00:45:51
    to serve EUDR compliance.
  • 00:45:53
    But to Tasso's point, the big picture
  • 00:45:55
    is deforestation, globally, and the climate crisis.
  • 00:45:58
    And so we will still need global data layers.
  • 00:46:01
    We will still need people actually monitoring
  • 00:46:03
    and reporting on how this whole EUDR ecosystem plays out
  • 00:46:07
    in the big picture.
  • 00:46:08
    Is there just displacement?
  • 00:46:09
    Or is there sort of massive economic failure
  • 00:46:12
    in certain places that actually increases famines or something
  • 00:46:15
    like that?
  • 00:46:15
    There's a bigger picture out there.
  • 00:46:16
    And EUDR is part of a system that's changing.
  • 00:46:19
    And so I think, on our picture, we would say,
  • 00:46:21
    we still monitor the globe.
  • 00:46:22
    We still build these global data layers that
  • 00:46:24
    are meant to be open, transparent,
  • 00:46:25
    and help scientists, help policymakers,
  • 00:46:27
    help journalists understand how the world is changing.
  • 00:46:30
    Specifically on the point around commercial data providers,
  • 00:46:32
    I think it's incredible to harness
  • 00:46:35
    the innovation of the private sector working
  • 00:46:37
    toward this problem.
  • 00:46:38
    And I think we see more and more of that every day.
  • 00:46:40
    I think we also see a lot of folks who do not
  • 00:46:42
    stop to document their work.
  • 00:46:43
    And if they did stop to document their work,
  • 00:46:46
    I think a lot of the buyers would be less than excited
  • 00:46:48
    about the results of that documentation.
  • 00:46:50
    And so I would encourage everyone, who is,
  • 00:46:52
    of course, a scrupulous actor and doing amazing stuff,
  • 00:46:55
    document it even if it's only internal
  • 00:46:56
    and your legal team won't let you release it.
  • 00:46:58
    But write that down.
  • 00:47:00
    What are these definitions?
  • 00:47:02
    What is this data?
  • 00:47:02
    How have we tested it?
  • 00:47:04
    What are the accumetrics we're reporting?
  • 00:47:07
    Have we almost red-teamed that process?
  • 00:47:10
    And even it's just for you, in your soul,
  • 00:47:12
    to know that you're selling something good,
  • 00:47:14
    it's important to have that documented.
  • 00:47:15
    And then it would be amazing if more people released these,
  • 00:47:18
    either through publication or simply through greater
  • 00:47:20
    transparency.
  • 00:47:21
    And I think, in our experience, as a nonprofit offering
  • 00:47:24
    products in this market, we get people
  • 00:47:27
    coming to us because of that, because of the transparency.
  • 00:47:30
    I don't think that will always be the case.
  • 00:47:32
    I hope that the bar will go up.
  • 00:47:33
    And we won't, maybe, in the future,
  • 00:47:35
    even need to offer products for companies in this way.
  • 00:47:38
    But people are coming to us because of the transparency,
  • 00:47:40
    because we can show everyone all of the code.
  • 00:47:42
    Because we can actually walk people through that.
  • 00:47:44
    And startups won't always be able to do that,
  • 00:47:46
    because you're competing for certain things.
  • 00:47:48
    But the more we can emulate that, as a community,
  • 00:47:51
    I think the better off we'll be.
  • 00:47:52
    ALICIA SULLIVAN: Awesome.
  • 00:47:53
    Does anyone else want to add?
  • 00:47:54
    PIERRICK RAMBAUD: Yeah.
  • 00:47:55
    Sure.
  • 00:47:55
    So I cannot agree more.
  • 00:47:58
    Please let me know what you're doing.
  • 00:47:59
    When you're a private provider, and you're coming to a company
  • 00:48:02
    and saying, we can manage your problem,
  • 00:48:04
    but you're not explaining your methodology,
  • 00:48:06
    we cannot really use your solution.
  • 00:48:08
    Because the issue that we have of definition,
  • 00:48:10
    the only solution that we can have
  • 00:48:12
    is transparency, transparency on your methodology.
  • 00:48:15
    Because if we can send your due diligence report and say,
  • 00:48:18
    this is the methodology that has been
  • 00:48:20
    used to perform the due diligence report, then at least
  • 00:48:23
    we are on the safe side when we are talking to the European
  • 00:48:26
    Commission.
  • 00:48:26
    If it's a black box, they have all the reason not
  • 00:48:29
    to trust you.
  • 00:48:29
    And we have all the reason not to work with you.
  • 00:48:33
    ALICIA SULLIVAN: Awesome.
  • 00:48:34
    I want to-- this is a great--
  • 00:48:35
    AUDIENCE: What does the regulation
  • 00:48:37
    say about the transparency?
  • 00:48:38
    ALICIA SULLIVAN: Yeah.
  • 00:48:39
    Thank you, Brady.
  • 00:48:40
    AUDIENCE: What does the regulation
  • 00:48:41
    say about transparency?
  • 00:48:42
    [LAUGHTER]
  • 00:48:44
    Say it.
  • 00:48:44
  • 00:48:47
    Like this is potentially a huge data
  • 00:48:49
    set for academics, watchdog groups, everything.
  • 00:48:52
    Will they have access to it or only
  • 00:48:54
    if there is actors that want to make it transparent?
  • 00:48:57
    Or is there something in the regulation about being open?
  • 00:48:59
    ALICIA SULLIVAN: Are you asking about
  • 00:49:01
    the specific due diligence reports or the data layers?
  • 00:49:04
    AUDIENCE: No, even the data that was going into that.
  • 00:49:06
    What is the--
  • 00:49:06
    ALICIA SULLIVAN: Remi, do you want to?
  • 00:49:07
    REMI D'ANNUNZIO: I'm happy to answer to that.
  • 00:49:09
    So the regulation says that all the information that
  • 00:49:11
    will be disclosed will be centralized
  • 00:49:13
    through an information system, which is not yet
  • 00:49:16
    completely defined.
  • 00:49:17
    That is supposed to come in December 2023.
  • 00:49:19
    So watch out.
  • 00:49:21
    [LAUGHTER]
  • 00:49:22
    And as per the--
  • 00:49:25
    I don't know, it's a principle of the EU.
  • 00:49:26
    All the data that will be inside this will be fully public.
  • 00:49:29
    It will be fully anonymized and fully public.
  • 00:49:31
    But that's going to be only the information that
  • 00:49:33
    is disclosed by the operators.
  • 00:49:34
    So the operators will disclose their fields
  • 00:49:38
    and the information from all those fields.
  • 00:49:42
    And it will be there with the associated information
  • 00:49:47
    on deforestation.
  • 00:49:49
    So this is going to be public.
  • 00:49:50
    And this is going to be fully out there.
  • 00:49:52
    AUDIENCE: So the polygons are available, but it's anonymized?
  • 00:49:54
    REMI D'ANNUNZIO: Yes.
  • 00:49:57
    That's what the regulation says.
  • 00:49:59
    ALICIA SULLIVAN: Well, and also, as part of the due diligence,
  • 00:50:01
    the operators, I believe, do have
  • 00:50:02
    to say what method they used, what data set, et cetera,
  • 00:50:05
    in order to prove their due diligence to the risk
  • 00:50:09
    assessment.
  • 00:50:10
    So I don't know that they have to actually submit
  • 00:50:12
    that particular data set, but they have to at least reference
  • 00:50:15
    it and how they got to that.
  • 00:50:17
    REMI D'ANNUNZIO: If I understand correctly,
  • 00:50:19
    the thing about the polygons is that you
  • 00:50:20
    have 100 polygons related to whatever is a company.
  • 00:50:23
    What you will not see is who is the holder of that polygon.
  • 00:50:28
    ALICIA SULLIVAN: The operator.
  • 00:50:29
    TASSO AZEVEDO: You just know that you have them
  • 00:50:32
    as a reference from something that comes.
  • 00:50:34
    So I think that's the anonymization
  • 00:50:36
    that we are talking about.
  • 00:50:37
    REMI D'ANNUNZIO: Yes, that's it.
  • 00:50:40
    So the reports will be--
  • 00:50:44
    I don't know if all the reports will be available, but yeah,
  • 00:50:47
    probably.
  • 00:50:48
    But the database, itself, will just
  • 00:50:50
    contain those geometries and the associated
  • 00:50:53
    risk of deforestation.
  • 00:50:55
    ALICIA SULLIVAN: Gotcha.
  • 00:50:56
    BOBBY PINTER: Good to know though, as well,
  • 00:50:58
    that some of the due diligence statements
  • 00:51:00
    will have to reference other due diligence statements.
  • 00:51:02
    So you'll get this kind of cascading set of due diligence
  • 00:51:05
    statement problem, which I think a lot of the work that we'll
  • 00:51:09
    end up doing will be actually rather
  • 00:51:11
    operational and procedural.
  • 00:51:13
    In my place in the market, it's going
  • 00:51:14
    to be, like, I don't hope to make any model layers, myself,
  • 00:51:17
    personally.
  • 00:51:18
    I would rather that I'm pulling from one
  • 00:51:20
    that's well peer-reviewed, that's recognized.
  • 00:51:22
    And there's going to be a lot of work on what
  • 00:51:24
    was the process we followed.
  • 00:51:26
    Because if it turns out that there's no deforestation,
  • 00:51:28
    it's fairly simple.
  • 00:51:29
    But if you're not sure or if there's a high enough risk,
  • 00:51:32
    like, oh, boy.
  • 00:51:33
    The process that you're going to have to go through?
  • 00:51:36
    I think there's going to be a lot of strange cost benefit
  • 00:51:37
    analysis.
  • 00:51:38
    And I think there's some points around exclusion
  • 00:51:40
    and segregation that were mentioned, as well.
  • 00:51:42
    Like what's it going to be worth to go and figure out
  • 00:51:45
    is there deforestation in this area?
  • 00:51:47
    You may reach a point where you're saying, well,
  • 00:51:49
    we're only sourcing a couple hundred pounds of product.
  • 00:51:52
    Am I going to go pay for a satellite tasking?
  • 00:51:55
    There's going to be these weird numbers
  • 00:51:57
    that organizations get to.
  • 00:51:59
    And to the point earlier, there's
  • 00:52:01
    a lot of economic ramifications.
  • 00:52:03
    I think part of where this ecosystem should go
  • 00:52:05
    is, like, I hope to not see myself
  • 00:52:06
    as being in the compliance industry.
  • 00:52:09
    I think there's a hope, here, that it's
  • 00:52:10
    going to be a little bit more of how are we both reshaping
  • 00:52:14
    supply chains, in a positive sense,
  • 00:52:15
    but also thinking a lot about what are the underlying
  • 00:52:19
    economics of these areas.
  • 00:52:21
    By moving supply chains away, what does that actually mean?
  • 00:52:25
    I'd like to see private actors and companies
  • 00:52:29
    taking some of that intentionality
  • 00:52:31
    into that approach.
  • 00:52:32
    I think a lot about, for example,
  • 00:52:33
    the stranded asset problem, where someone's, like,
  • 00:52:35
    oh, we sold away a particularly expensive or carbon-intensive
  • 00:52:41
    part of our business.
  • 00:52:42
    Someone still bought it.
  • 00:52:43
    It didn't really-- it didn't poof out of existence, right?
  • 00:52:46
    And I think those are the kinds of activities
  • 00:52:48
    that it's probably on this community, in a certain sense,
  • 00:52:51
    to say, how do we make sure that we just don't have
  • 00:52:53
    some super companies that can just exclude parts
  • 00:52:56
    of the market or change their supply chains in a way that
  • 00:52:58
    leave a lot of groups behind.
  • 00:52:59
    ALICIA SULLIVAN: All right, guys,
  • 00:53:01
    we have five minutes left.
  • 00:53:02
    I know we have at least two more questions.
  • 00:53:04
    Brady?
  • 00:53:05
    AUDIENCE: Yeah.
  • 00:53:05
    Hi.
  • 00:53:06
    I was wondering-- I have two questions.
  • 00:53:08
    One was, when you're talking about deforestation risk,
  • 00:53:11
    is it at the point of planting, is it during the process,
  • 00:53:15
    is it at sowing?
  • 00:53:16
    Are you calculating the risk of deforestation
  • 00:53:19
    even before any of these processes actually start?
  • 00:53:23
    And at what resolution are you measuring?
  • 00:53:26
    How do you say that a 3 meter or 50 centimeter
  • 00:53:29
    resolution map of deforestation, and some other company does
  • 00:53:33
    a coarser resolution versus a finer resolution--
  • 00:53:35
    how do you even say this better than this?
  • 00:53:39
    Like is there even an uncertainty marking in any way?
  • 00:53:42
    And the next question I have is, is there
  • 00:53:44
    any concept of something like blockchain monitoring of food
  • 00:53:48
    traceability?
  • 00:53:49
    People have tried this approach, where, at every step,
  • 00:53:53
    there is a monitoring that goes in,
  • 00:53:55
    in terms of auditors, in terms of where the food actually
  • 00:53:58
    comes from, apart from doing high level risk
  • 00:54:02
    assessment from satellites.
  • 00:54:03
    I was just curious.
  • 00:54:04
  • 00:54:07
    REMI D'ANNUNZIO: OK.
  • 00:54:08
    On the information that needs to come?
  • 00:54:11
    At the moment, the only real information
  • 00:54:14
    was, the 31st of December 2020, was your field forest or not?
  • 00:54:19
    That's the only information you actually need to provide.
  • 00:54:22
    Because you say, hey, this is my field.
  • 00:54:25
    Here, is where I grow my crop.
  • 00:54:27
    So if this is where my crop is and this was forest in 2020,
  • 00:54:30
    I deforested-- end of the story.
  • 00:54:33
    There is no real specification.
  • 00:54:36
    There's specification on the field precisions.
  • 00:54:39
    And I think it's something around 3 meters.
  • 00:54:42
    But there are no other specifications
  • 00:54:44
    on what type of data, on what you need to do.
  • 00:54:47
    It's just like you have to make a due diligence on whether you
  • 00:54:51
    deforested or not.
  • 00:54:52
    And in the end, it boils down to, was it forest in 2020.
  • 00:54:56
    If it was, it's untouchable forever.
  • 00:54:59
    And if it was not forest in 2020, there is still somehow,
  • 00:55:03
    in 2025 maybe in 2026, where we're going to have regrowth,
  • 00:55:06
    you're going to still have to prove
  • 00:55:07
    that it was not deforested.
  • 00:55:09
    So it didn't grow into a forest.
  • 00:55:11
    But that's the only thing that's interesting.
  • 00:55:13
    So we don't even actually need information on deforestation.
  • 00:55:15
    We need information on status of forest,
  • 00:55:17
    which is why it's so important to have
  • 00:55:19
    the difference between crops that
  • 00:55:23
    look like forests and forest.
  • 00:55:26
    AUDIENCE: But it's not continuous monitoring.
  • 00:55:28
    So you don't have to-- you just have to say,
  • 00:55:30
    2020, it was not forest.
  • 00:55:32
    And we're good for the next 10 years.
  • 00:55:34
    We don't have to report on this every year or every month
  • 00:55:37
    or something?
  • 00:55:37
    REMI D'ANNUNZIO: Unclear.
  • 00:55:37
    Unclear.
  • 00:55:38
    PIERRICK RAMBAUD: You will actually
  • 00:55:39
    need to report if the plot has changed crop.
  • 00:55:43
    AUDIENCE: OK.
  • 00:55:43
    PIERRICK RAMBAUD: So that's the first thing.
  • 00:55:45
    Then, when you were asking, when should we
  • 00:55:47
    check if it's actually deforested or not,
  • 00:55:50
    don't forget that there's cattle in the list.
  • 00:55:52
    And they are actually born, grown,
  • 00:55:54
    and killed in different places.
  • 00:55:56
    So plants are point of growth and harvesting,
  • 00:56:00
    but, for cattle, that's a bit different.
  • 00:56:01
    And when you are speaking about definition,
  • 00:56:05
    so there's definition on the geometry.
  • 00:56:08
    So it needs to use six digits for the GPS coordinates.
  • 00:56:12
    But there is no other specification.
  • 00:56:14
    So if you want to do a triangle, with three points,
  • 00:56:16
    with six digits, apparently it works.
  • 00:56:19
    ALICIA SULLIVAN: I think there was one more question.
  • 00:56:22
    And we only have two minutes.
  • 00:56:23
    So I wanted to get to that real quick.
  • 00:56:25
    AUDIENCE: I just wanted to ask about--
  • 00:56:27
    you spoke about anonymizing polygons.
  • 00:56:30
    And I'm thinking, well, you can't really anonymize it.
  • 00:56:33
    If it's got a lat and long, you know where it is.
  • 00:56:36
    And the farmers might not want to tell you
  • 00:56:40
    where their plots are.
  • 00:56:42
    And how do you deal with those sorts of issues?
  • 00:56:46
    TASSO AZEVEDO: Yeah, if they can't tell,
  • 00:56:47
    they will not be able to sell, obviously.
  • 00:56:49
    But I think the question is like it simply
  • 00:56:52
    makes absolutely no sense that this is done by plot.
  • 00:56:56
    That's not how deforestation happens.
  • 00:56:58
    If we go, deforestation is growing every year.
  • 00:57:00
    So if you take a typical farm in Mato Grosso, for example,
  • 00:57:04
    for soybean, you have 200 hectares of one plot.
  • 00:57:08
    100 hectares were deforested before 2020
  • 00:57:13
    another one was after.
  • 00:57:15
    And then this is the plot.
  • 00:57:16
    What do you expect that he goes and say,
  • 00:57:18
    oh, now I have two plots.
  • 00:57:19
    This plot, here, I can sell to Europe.
  • 00:57:23
    This other plot, I can't sell to Europe.
  • 00:57:24
    It's completely ridiculous.
  • 00:57:26
    It makes absolutely no sense to make that.
  • 00:57:28
    So it's kind of--
  • 00:57:29
    [LAUGHTER]
  • 00:57:29
    I think that those things--
  • 00:57:31
    no, it's very simple.
  • 00:57:32
    It's very a practical topic.
  • 00:57:35
    There is no way that you will be able to separate it.
  • 00:57:37
    No way.
  • 00:57:38
    There is no way.
  • 00:57:38
    Nobody will harvest things and separate in two trucks
  • 00:57:41
    and things like this.
  • 00:57:42
  • 00:57:44
    In practical terms, what will happen is
  • 00:57:46
    you have to work in properties.
  • 00:57:48
    Otherwise you always run on the risk
  • 00:57:50
    that you have this contamination and so on.
  • 00:57:52
    REMI D'ANNUNZIO: But I think this is
  • 00:57:53
    why they make it complicated.
  • 00:57:54
    I think their initial goal was to actually prevent
  • 00:57:57
    deforestation to happen.
  • 00:57:58
    And it's so complicated.
  • 00:57:59
    As Bobby was saying, if you don't deforest,
  • 00:58:01
    if you don't have deforestation, here and then here,
  • 00:58:04
    then you're fine.
  • 00:58:05
    And this is basically supply chains
  • 00:58:07
    that don't have deforestation will be good to go.
  • 00:58:11
    Supply chains that are associated with deforestation
  • 00:58:14
    will not enter the market, because it's so complex,
  • 00:58:16
    and it's so risky that they will not do it.
  • 00:58:19
    If there is any risk--
  • 00:58:20
    there was one thing I wanted to finish on.
  • 00:58:22
    If there's any risk of suspicion on any part of the shipments,
  • 00:58:25
    it says that the country receiving
  • 00:58:27
    have the right to stop the containers for three days.
  • 00:58:30
    If we think about how much it's going
  • 00:58:33
    to cost to have ships waiting for verification to come?
  • 00:58:38
    And if there is no verification, then they
  • 00:58:40
    have the right to prolong.
  • 00:58:41
    It's a deterring mechanism.
  • 00:58:43
    ALICIA SULLIVAN: Yeah.
  • 00:58:44
    No, it definitely is.
  • 00:58:45
    And we are out of time.
  • 00:58:46
    But I can tell that this is a very hot topic.
  • 00:58:48
    [LAUGHTER]
  • 00:58:49
    I am very glad that we had this amazing discussion.
  • 00:58:53
    So I'd encourage you all, if you're
  • 00:58:55
    interested in getting into the next level tech session, about
  • 00:58:59
    commodity mapping and understanding
  • 00:59:00
    the current state of that, that is happening at 3 o'clock
  • 00:59:03
    in the same room.
  • 00:59:05
    And please give a big round of applause to our panelists.
  • 00:59:08
    [APPLAUSE]
  • 00:59:10
    [MUSIC PLAYING]
  • 00:59:15
Tags
  • EUDR
  • deforestation
  • supply chain
  • geospatial data
  • compliance
  • regulation
  • sustainability
  • traceability
  • commodities
  • deforestation-free